An Alternative Way to install Juno
Julia is a high-performance and dynamic programming language that was developed in 2012 by Jeff Bezanson, Stefan Karpinski, and Viral B. Shah. It is open-source and fast-emerging. It was specifically designed to quickly implement basic mathematics and originally built for data science, particularly, machine learning. As the saying goes, “Walks like Python, runs like C”, it is built majorly on C.
As other languages have IDEs (Integrated Development Environments) for writing, running, and building projects, there are various IDEs for Julia. Some of the well known are Juno, Weave, Spyder, LightHouse. Also, Visual Studio Code has an extension for the language. The most used IDE is Juno.
Juno is an IDE built for Julia which has all the necessary features used for development in the language. To install Juno, there are two requirements; Julia and Atom. This is because Juno is built on Atom.T he minimum requirement needed to install Juno the ideal way is having at least version 1.0 of Julia and the latest version of Atom installed, then in Atom, the package “uber-juno” is downloaded and installed. Some of the time though because of some dependency issues as the package is still in development, it doesn’t work and I would be sharing one of the best alternatives to have Juno on Atom.
This method can be used when the previous method of installation has issues. The tools/packages needed for this mode of installation are Nodejs, Python 2.7, npm, apm, and apx.
- Install Julia
When installing Julia for the first time, the error of “Error: Installing “julia-client@0.8.10” failed” may come up. Julia-client is the core package of Juno which if not installed will not allow Juno to run on Atom. The other method to install julia-client is to first install Nodejs, then open Powershell as Administrator and run the following commands, `npm install — global — production windows-build-tools` which would take a while (around 15 minutes). This would install some Microsoft Visual Studio Tools and Python 2.7 (I would also share how to remove Python 2.7 from path after installing this incase another version of Python is installed on the system). Then the second command `npm install — global node-gyp` is run. This should install julia-client on your system.
2. Reinstall uber-juno
After julia-client has been installed, uninstall uber-juno then install it again. This should install Juno on Atom. Then to verify it is working perfectly on Atom, start a session and try running the REPL. The issue “julia not added to path” may come up. To add Julia to path, the path “C:\Users\YourUserName\AppData\Local\Programs\Julia\Julia-1.4.1\bin” should be placed in the “Julia Path” option.
3. Install dependencies
After this Juno might not still run. This is because another build/dependency is missing. The name of this dependency is ink. To install ink, the command `apx install ink` is used. Apx is an alternative package manager for Atom. All about apx is at https://github.com/Aerijo/apx. Then run the command `apx install ink` again. After this, both ink and julia-client are installed.
4. Uninstall and re-install uber-juno
Finally, uninstall and re-install uber-juno. Then automatically, Juno will start running and another error `Uncaught TypeError: Cannot read property ‘close’ of undefined` will appear. This can be fixed in two ways: Either by restarting Atom or by uninstalling julia client and ink then reinstalling it again. First try restarting atom. If that doesn’t work, then you can follow the installation guide `apm uninstall ink`, `apm uninstall julia-client`, `apm install ink`, and `apm install julia-client`.
Now, Juno has been successfully installed on Atom and you can start the REPL and run any .jl file. If you have any version of Python installed and want to remove Python 2.7 from path, open up Control Panel\Programs\Uninstall a Program and click on the Python 2.7.5 (64-bit) and click on the “Change” option. Then the Python Application Wizard comes up and then click on “Remove Python 2.7.5 (64-bit)”. This would remove Python 2.7 from path but not uninstall it. Now your other version of Python is now on path again.
I hope this solved your problem, if it didn’t, you can reach out to me on victoriaolusheye@gmail.com and we can work something out together as this is an alternative.
Thanks!
